Update README.md

README.md

@@ -22,6 +22,63 @@ CrossLing-OCR-Mini is optimized for **low-resource and structurally complex lang
   
 Experimental results show that CrossLing-OCR-Mini **outperforms or matches mainstream OCR systems** on multiple low-resource languages.
 
+
+## 🚀 Usage / Inference
+
+You can easily perform inference with CrossLing-OCR-Mini using the 🤗 Transformers library.
+The following example demonstrates a simple OCR inference pipeline on a single image.
+
+🔧 Requirements
+
+Python ≥ 3.8
+
+transformers (latest recommended)
+
+CUDA-enabled GPU (recommended for better performance)
+
+```
+pip install -U transformers accelerate
+```
+
+## 🧪 Simple OCR Inference Example
+```
+from transformers import AutoModel, AutoTokenizer
+import os
+
+# Path or Hugging Face model id
+model_id = "checkpoint-80000-merged"
+
+# Load tokenizer and model
+tokenizer = AutoTokenizer.from_pretrained(
+    model_id,
+    trust_remote_code=True
+)
+
+model = AutoModel.from_pretrained(
+    model_id,
+    trust_remote_code=True,
+    low_cpu_mem_usage=True,
+    device_map="cuda",
+    use_safetensors=True,
+    pad_token_id=tokenizer.eos_token_id
+)
+
+model = model.eval().cuda()
+
+# Input image for OCR
+image_file = "test.png"
+
+# Perform plain text OCR
+result = model.chat(
+    tokenizer,
+    image_file,
+    ocr_type="ocr"
+)
+
+print("Predicted OCR result:\n")
+print(result)
+```
+
 ---
 
 ## 🧪 Performance Notes & Limitations